이메일 #026 - #030

원본 출처: Satoshi - Sirius emails 2009-2011

이메일 #026

Date: Wed, 29 Jul 2009 18:14:51 +0300
From: [email protected]
To: Satoshi Nakamoto <[email protected]>
Subject: Re: Bitcoin

영어 원문

I've had quite a few errors coming up when trying to build the  
third-party libraries and adding them to the Bitcoin build. Do you  
happen to have a ready-to-build package that you could upload to the  
CVS or somewhere else? I use mingw + msys, but I guess I could try  
Visual C++ also, if it's easier that way.

이메일 #027

Date: Mon, 24 Aug 2009 06:38:13 +0300
From: [email protected]
To: Satoshi Nakamoto <[email protected]>
Subject: Re: Bitcoin

영어 원문

I got it compile with MinGW + MSYS when I used wxPack instead of just  
wxWidgets. Maybe wxAdditions was required. The bitcoin.exe filesize  
was 52MB though, I should see how that can be fixed.

Next I'm going to implement the "minimize to tray" feature and the  
option to autostart Bitcoin with Windows, so the number of nodes  
online would stay higher. After that I could see if I can do a Linux  
port or the command line interface needed for web app frameworks.

Drop by at #bitcoin-dev on FreeNode some time if you use IRC.

And again, thanks for the great work you've done with Bitcoin.

Quote [email protected]:

> I've had quite a few errors coming up when trying to build the
> third-party libraries and adding them to the Bitcoin build. Do you
> happen to have a ready-to-build package that you could upload to the
> CVS or somewhere else? I use mingw + msys, but I guess I could try
> Visual C++ also, if it's easier that way.

이메일 #028

Date: Mon, 24 Aug 2009 23:00:35 +0100
From: Satoshi Nakamoto <[email protected]>
Subject: Re: Bitcoin
To: [email protected]

영어 원문

That's a good point that since you know how many coins exist and how 
fast new ones are created, you could set a support price based on the 
amount of legacy currency you have and be sure you'll have enough to 
meet all demands.  I had imagined an auction, but it would be far 
simpler and more confidence inspiring to back it at a specific exchange 
rate.

Offering currency to back bitcoins would attract freebie seekers, with 
the benefit of attracting a lot of publicity.  At first it would mostly 
be seen as a way to get free money for your computer's idle time.  Maybe 
pitched like help support the future of e-commerce and get a little 
money for your computer's spare cycles.  As people cash in and actually 
get paid, word would spread exponentially.

It might help to keep the minimum transaction size above an amount which 
a typical user would be able to accumulate with one computer, so that 
users have to trade with each other for someone to collect enough to 
cash in.  Aggregators would set up shop to buy bitcoins in smaller 
increments, which would add confidence in users ability to sell bitcoins 
if there are more available buyers than just you.

People would obviously be sceptical at first that the backing will hold 
up against an onslaught of people trying to get the free money, but as 
the competition raises the proof-of-work difficulty, it should become 
clear that bitcoins stay scarce.  People will see that they can't just 
get all the bitcoins they want.  It would establish a minimum value 
under bitcoins enabling them to be used for other purposes if, 
hopefully, other purposes are waiting for something to use.

>> It would help if there was something for people to use it for.  We need
>> an application to bootstrap it.  Any ideas?
> 
> I've been thinking about a currency exchange service that sells and
> buys bitcoins for euros and other currencies. Direct exchangeability
> to an existing currency would give bitcoin the best possible initial
> liquidity and thus the best adoptability for new users. Everyone
> accepts payment in coins that are easily exchangeable for common
> money, but not everyone accepts payment in coins that are only
> guaranteed to buy a specific kind of a product.

That would be more powerful if there was also some narrow product market 
to use it for.  Some virtual currencies like Tencent's Q coin have made 
headway with virtual goods.  It would be sweet if there was some way to 
horn in on a market like that as the official virtual currency gets 
clamped down on with limitations.  Not saying it can't work without 
something, but a ready specific transaction need that it fills would 
increase the certainty of success.

> At its simplest this exchange service could be a website where
> traders, who can be individual persons, can post their rates, and
> random users can leave trade requests. Some kind of an average rate
> estimate could be shown on the site. Small-scale trading by
> individuals would be outside legal hassle in most countries, and
> putting all the eggs in the same basket would be avoided.

Basically like an eBay site with user reviews to try to establish which 
sellers can be trusted.  The escrow feature will help but not solve 
everything.  It would be far more work to set up such a site than just 
to set up a single exchange site of your own, and there won't be enough 
users to make it go until later.  I'm thinking it wouldn't make sense to 
make an eBay type site until later.

> Another idea, which could be additional to the previous one, would be
> an automated exchange service. The service would automatically
> calculate the exchange rate and perform the transactions. This would
> be nicer to the user: completion of the transaction request would be
> certain and instantaneous. Making this service might actually be quite
> easy if there was a command line interface to Bitcoin: just take any
> web application framework and use PayPal back-end integration to
> automatically send euros when Bitcoins are received, and vice versa.
> This kind of business would also work great on larger scale if you set
> up a company and take care of all the bureaucracy needed to practice
> currency exchange. (I actually have a registered company that I've
> used for billing of some IT work, I could use that as a base.)

Even if you had automation, you'd probably want to review orders 
manually before processing them anyway.  It wouldn't be hard to process 
orders by hand, especially at first.  You could always set a minimum 
order size to keep orders more infrequent.

> This exchange business thing is something that I'd be interested in
> doing, and I also have the sufficient technical skills to do it.
> Although, before this can be done, there should be a non-alpha version
> of Bitcoin (and the command line interface / API).
> 
> If this gets started, donors / high-risk investors would be very
> welcome to bring capital for the currency's backup.
> 
> So, what do you think about the idea? Note that this is not something
> that I'm asking you to do (unless you want to) if you're busy with
> other things. I can do it myself, if I get positive reviews about the
> plan.

That's great, I could probably get a donor to send currency to you which 
you convert to euros and pay out through methods that are convenient for 
users.  I don't want to do an exchange business myself, but it can be 
done independently of me.  Like you say, there is more software 
development to be done first, and also I'd like to keep trying for a 
while to think of a bootstrap application to use bitcoins for.  I've had 
some ideas that could only be done before an exchange exists.

BTW, I tried to buy bitcoin.com before I started but there was no 
chance, it's owned by a professional domain speculator.  It's normal for 
open source projects to have .org so it's not so bad.

이메일 #029

Date: Mon, 24 Aug 2009 23:04:25 +0100
From: Satoshi Nakamoto <[email protected]>
Subject: Re: Bitcoin
To: [email protected]

영어 원문

Glad that worked, it's a pain that the dependencies are so big and hard 
to build.  Some of them give little attention to the Windows build. 
Next time I update to the latest versions, maybe I'll lay everything out 
in one directory tree and bundle the whole thing up into a giant archive.

I'm not sure they had wxPack before.  I'm glad they got that so everyone 
doesn't have to build wxWidgets themselves.  OpenSSL is the harder one 
to build.

I reduced the EXE size by running strip.exe on it to take out the debug 
symbols.  That's with mingw.  That's the better compiler, I only used VC 
  for debugging.

[email protected] wrote:
> I got it compile with MinGW + MSYS when I used wxPack instead of just 
> wxWidgets. Maybe wxAdditions was required. The bitcoin.exe filesize was 
> 52MB though, I should see how that can be fixed.
> 
> Next I'm going to implement the "minimize to tray" feature and the 
> option to autostart Bitcoin with Windows, so the number of nodes online 
> would stay higher. After that I could see if I can do a Linux port or 
> the command line interface needed for web app frameworks.
> 
> Drop by at #bitcoin-dev on FreeNode some time if you use IRC.
> 
> And again, thanks for the great work you've done with Bitcoin.
> 
> Quote [email protected]:
> 
>> I've had quite a few errors coming up when trying to build the
>> third-party libraries and adding them to the Bitcoin build. Do you
>> happen to have a ready-to-build package that you could upload to the
>> CVS or somewhere else? I use mingw + msys, but I guess I could try
>> Visual C++ also, if it's easier that way.
> 
> 
>

이메일 #030

Date: Fri, 28 Aug 2009 07:10:06 +0300
From: [email protected]
To: Satoshi Nakamoto <[email protected]>
Subject: Re: Bitcoin

영어 원문

> It might help to keep the minimum transaction size above an amount
> which a typical user would be able to accumulate with one computer, so
> that users have to trade with each other for someone to collect enough
> to cash in.  Aggregators would set up shop to buy bitcoins in smaller
> increments, which would add confidence in users ability to sell
> bitcoins if there are more available buyers than just you.

That might be a good idea.

> That would be more powerful if there was also some narrow product
> market to use it for.  Some virtual currencies like Tencent's Q coin
> have made headway with virtual goods.  It would be sweet if there was
> some way to horn in on a market like that as the official virtual
> currency gets clamped down on with limitations.  Not saying it can't
> work without something, but a ready specific transaction need that it
> fills would increase the certainty of success.

Bitcoin could be promoted to the users of virtual communities like  
World of Warcraft and Second Life, which both have millions of users.  
It would be great if not only peer-to-peer item traders, but also  
providers of some existing virtual services that already have a lot of  
customers, were to adopt the currency early on.

A programming question: What do you think about using the Boost's  
program_options to write settings like the transaction fee into a file  
bitcoin.config? Or is it better to save them in the database as it is  
now? Having a config file would make it easier to change the settings  
when running the program on a remote server with a console access only.